
Design a safety checklist for AI use
You'll have a one-page visual checklist (infographic or PDF template) showing red flags and approval steps for when you hand off AI work to others or deploy it in your organization.
What you'll make
AI Deployment Safety Checklist
- Header: The deployment decision tree
- Left panel: Red flags that require human oversight
- Right panel: Approval gates by consequence level
- Footer: When to pause and escalate
Upload your final checklist as a PDF, infographic, or image file that someone could print and use immediately
The 3 parts
Each part ends with a check you can see.
Map Your AI Deployment Patterns
Identify how AI is currently being used in your organization and what could go wrong.
Done when
You have 5-8 current AI use cases documented with their potential failure modes and current oversight levels.
Find Your Oversight Boundaries
Test AI on borderline cases to discover where human oversight becomes mandatory.
Done when
You have tested 3 AI use cases at different oversight levels and documented which required human intervention to prevent problems.
Design Your Visual Safety Checklist
Build a one-page checklist that translates your findings into deployment decisions anyone can follow.
Done when
You have a visual checklist with specific red flags, approval gates, and escalation triggers that someone else could use to evaluate AI deployments.
Capstone: you put together your AI Deployment Safety Checklist.
Where this helps
An elective: a single lesson that sits outside rtwo's job certificates and doesn't count toward one.
About this lesson
What you'll be able to do
- Create a visual safety checklist that stops dangerous AI deployments before they happen
- Test the checklist on 3 real AI use cases in your organization
- Know when to require human oversight vs when AI can run unsupervised
- Have specific red flags that trigger immediate AI output review
Who it's for
- Team leads who need to approve AI tools for their people
- Project managers rolling out AI workflows
- Anyone who's been asked 'is it safe to use AI for X?' and wants a systematic answer
What it won't cover
- Technical AI safety research or alignment theory
- Legal compliance frameworks (SOX, GDPR, etc.)
- Building AI systems from scratch — this is about using existing tools safely
Skills you'll practise
- Risk pattern recognition in AI outputs
- Deployment boundary setting
- Stakeholder communication about AI limits
- Systematic AI output validation